How is Emotional Evidence from Multiple Sources Used in Perceptual Decision Making?

Abstract

Judging the overall emotional nature of a scene often requires us to deliberately integrate pieces of evidence that range from neutral to emotional in nature. Research so far has focused on relatively rapid decisions. Hence, how evidence is valuated and sampled given ample time, and the precise computations involved in integrating emotion-related evidence to form an overall perceptual decision remain unclear. Further, it is unclear how this process is influenced by decision-makers’ endogenous (voluntary) attention directed towards emotional evidence. In two experiments, human subjects (N=33 and N = 40) viewed arrays of faces with expressions ranging from extremely fearful to extremely happy. They decided whether the faces were more fearful or happier on average in the absence (Experiment 1) and presence (Experiment 2) of endogenous attention directed towards the fearful or happy faces. Results showed that subjects evaluated the same emotional faces differently depending on the face array context. While they attentionally sampled extreme emotional evidence more, they did not weigh it higher or lower in their decision making. Rather, their decision-making strategy was best modeled by weighing all pieces of context-dependent subjective evaluated emotional evidence equally. Furthermore, endogenous attention influenced sampling of the emotional evidence, but the decision-making strategy remained unchanged. In conclusion, the findings show that when forming a judgment about multiple emotional stimuli without time pressure, humans weigh multiple pieces of subjectively valuated context-dependent evidence equally. These findings elucidate specific computational mechanisms of emotion-related decision making, shedding light on the mechanisms of normal and pathological affective decisions.